RE: Deployment Failures and Memcache Unavailability Due to Underlying Component

[Nickolas Daskalou]

For anyone else having this issue....

We found that during this time, even if a deployment succeeds, the underlying App Engine dispatcher could also "break" during a deployment.

The only way to fix it was to push our existing dispatch rules again (for Python standard: "appcfg.py update_dispatch .").

Hope that helps others and also gives the Google Cloud engineering/monitoring teams clues on how to fix it.

Nick

[Nicholas (Google Cloud Support)]

Hey Nickolas,

Thank you for sharing your findings with the community here in order to help!  This is much appreciated.  Nonetheless, if I've understood your message correctly, you've occasionally encountered errors with appcfg.py where deployment which includes your dispatch.yaml does not update your dispatch rules despite successfully updating your App Engine service.  This is unexpected behavior.  As such, I'd recommend reporting this on our Issue Tracker and including any information that might be useful such as:

• Time frame for when this happened
  
• The runtime being used when deploying the App Engine service
  
• The dispatch.yaml being uploaded during the service update
  

If you choose to file a new issue, feel free to post a link to the issue here so that others in the community can follow along.  In the meantime, I would consider using gcloud app deploy to deploy App Engine services and explicitly providing your dispatch.yaml as an argument.